Carrier aggregation (CA), including intra-node and inter-node Carrier Aggregation, was introduced in Release 10 of the Evolved Universal Terrestrial Network (E-UTRAN) standard as a means for qualifying E-UTRAN to meet the requirements for 4G (1000 Mbit/s) as well as for allowing operators with small (less than 20 MHz) scattered spectrum allocations to provide a good user experience by aggregating the scattered allocations into, e.g., bandwidths of 10 MHz, 20 MHz or more.
The user equipment (UE) is connected to a serving cell termed Primary Cell (PCell) on what is referred to as the Primary Component Carrier (PCC). Mobility is catered for on this carrier. In case the UE is using services that require high throughput, the network may activate one or more additional serving cells, each termed Secondary Cell (SCell), on what is referred to as Secondary Component Carrier(s) (SCC(s)). The activation may happen before or after the SCell has been detected by the UE.
Two types of aggregation scenarios are considered for Release 10 (e.g., according to documents 3GPP TS 36.104 V10.11.0 and 3GPP TR 36.808 V10.1.0): (i) intra-band contiguous aggregation, and (ii) inter-band aggregation. In Release 11 (e.g., according to document 3GPP TS 36.104 V11.6.0, Sect. 6.5.3.1 and Table 5.5-4), one more is considered: (iii) intra-band non-contiguous aggregation.
For intra-band contiguous carrier aggregation, the PCell and one or more SCells are contiguous in frequency. It is required from the standard that for contiguous intra-band aggregation, the time difference between PCell and SCell is allowed to be at most ±130 ns (3GPP TS 36.104, V11.4.0, sub-clause 6.5.3). It is further assumed in the standard that for this particular scenario, one can use a single Fast Fourier Transform (FFT) operation or block to demodulate the signal from both PCell and SCell simultaneously. Thus, in practice, it is required that the PCell and SCell are collocated, i.e., transmitted from the same site, since otherwise propagation delay would make it impossible to use a single FFT.
For intra-band non-contiguous aggregation the timing difference is allowed to be at most ±260 ns, but no assumption is made on that the cells are co-located or that a single FFT can be used.
For inter-band carrier aggregation the timing difference between the PCell and SCell is allowed to be at most ±260 ns. However, for this scenario it is further assumed that the cells may be non-collocated and that the UE will have to cope with a propagation delay difference between PCell and SCell of up to ±30 μs, resulting in a maximum delay spread of ±30.26 μs (3GPP TS 36.300, V11.5.0, Annex J).
Existing techniques for aggregating radio resources are not suitable for the increasing complexity of cellular networks, e.g., heterogeneous networks including cells operated at largely different power levels, such as macro-cells and pico-cells. With inter-node radio resource aggregation, new deployment scenarios are encountered. Not all timings for cells under which the UE has coverage simultaneously are such that they fall within the time difference that the UE can handle, e.g. ±30.26 μs. Hence some cells will not be suitable to use for aggregation, but the UE has no means for indicating which cells can be used to the network.
Later network deployments (3GPP TS Rel. 12 and onwards, including 5G) will be even more complex, and it will be virtually impossible to predict and configure the coverage a UE is provided at any given location, or the timing difference the UE observes among all cells within the coverage the UE is currently located. To assess such information, extensive drive tests would be needed every time a new cell or frequency layer is added in the area.